##// END OF EJS Templates
revlog: change generaldelta delta parent heuristic...
revlog: change generaldelta delta parent heuristic The old generaldelta heuristic was "if p1 (or p2) was closer than the last full text, use it, otherwise use prev". This was problematic when a repo contained multiple branches that were very different. If commits to branch A were pushed, and the last full text was branch B, it would generate a fulltext. Then if branch B was pushed, it would generate another fulltext. The problem is that the last fulltext (and delta'ing against `prev` in general) has no correlation with the contents of the incoming revision, and therefore will always have degenerate cases. According to the blame, that algorithm was chosen to minimize the chain length. Since there is already code that protects against that (the delta-vs-fulltext code), and since it has been improved since the original generaldelta algorithm went in (2011), I believe the chain length criteria will still be preserved. The new algorithm always diffs against p1 (or p2 if it's closer), unless the resulting delta will fail the delta-vs-fulltext check, in which case we delta against prev. Some before and after stats on manifest.d size. internal large repo old heuristic - 2.0 GB new heuristic - 1.2 GB mozilla-central old heuristic - 242 MB new heuristic - 261 MB The regression in mozilla central is due to the new heuristic choosing p2r as the delta when it's closer to the tip. Switching the algorithm to always prefer p1r brings the size back down (242 MB). This is result of the way in which mozilla does merges and pushes, and the result could easily swing the other direction in other repos (depending on if they merge X into Y or Y into X), but will never be as degenerate as before. I future patch will address the regression by introducing an optional, even more aggressive delta heuristic which will knock the mozilla manifest size down dramatically.

File last commit:

r25984:c57509e8 default
r26117:4dc5b51f default
Show More
templatekw.py
500 lines | 16.8 KiB | text/x-python | PythonLexer
# templatekw.py - common changeset template keywords
#
# Copyright 2005-2009 Matt Mackall <mpm@selenic.com>
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2 or any later version.
from __future__ import absolute_import
from .node import hex
from . import (
error,
hbisect,
patch,
scmutil,
util,
)
# This helper class allows us to handle both:
# "{files}" (legacy command-line-specific list hack) and
# "{files % '{file}\n'}" (hgweb-style with inlining and function support)
# and to access raw values:
# "{ifcontains(file, files, ...)}", "{ifcontains(key, extras, ...)}"
# "{get(extras, key)}"
class _hybrid(object):
def __init__(self, gen, values, makemap, joinfmt=None):
self.gen = gen
self.values = values
self._makemap = makemap
if joinfmt:
self.joinfmt = joinfmt
else:
self.joinfmt = lambda x: x.values()[0]
def __iter__(self):
return self.gen
def __call__(self):
makemap = self._makemap
for x in self.values:
yield makemap(x)
def __contains__(self, x):
return x in self.values
def __len__(self):
return len(self.values)
def __getattr__(self, name):
if name != 'get':
raise AttributeError(name)
return getattr(self.values, name)
def showlist(name, values, plural=None, element=None, separator=' ', **args):
if not element:
element = name
f = _showlist(name, values, plural, separator, **args)
return _hybrid(f, values, lambda x: {element: x})
def _showlist(name, values, plural=None, separator=' ', **args):
'''expand set of values.
name is name of key in template map.
values is list of strings or dicts.
plural is plural of name, if not simply name + 's'.
separator is used to join values as a string
expansion works like this, given name 'foo'.
if values is empty, expand 'no_foos'.
if 'foo' not in template map, return values as a string,
joined by 'separator'.
expand 'start_foos'.
for each value, expand 'foo'. if 'last_foo' in template
map, expand it instead of 'foo' for last key.
expand 'end_foos'.
'''
templ = args['templ']
if plural:
names = plural
else: names = name + 's'
if not values:
noname = 'no_' + names
if noname in templ:
yield templ(noname, **args)
return
if name not in templ:
if isinstance(values[0], str):
yield separator.join(values)
else:
for v in values:
yield dict(v, **args)
return
startname = 'start_' + names
if startname in templ:
yield templ(startname, **args)
vargs = args.copy()
def one(v, tag=name):
try:
vargs.update(v)
except (AttributeError, ValueError):
try:
for a, b in v:
vargs[a] = b
except ValueError:
vargs[name] = v
return templ(tag, **vargs)
lastname = 'last_' + name
if lastname in templ:
last = values.pop()
else:
last = None
for v in values:
yield one(v)
if last is not None:
yield one(last, tag=lastname)
endname = 'end_' + names
if endname in templ:
yield templ(endname, **args)
def getfiles(repo, ctx, revcache):
if 'files' not in revcache:
revcache['files'] = repo.status(ctx.p1(), ctx)[:3]
return revcache['files']
def getlatesttags(repo, ctx, cache):
'''return date, distance and name for the latest tag of rev'''
if 'latesttags' not in cache:
# Cache mapping from rev to a tuple with tag date, tag
# distance and tag name
cache['latesttags'] = {-1: (0, 0, ['null'])}
latesttags = cache['latesttags']
rev = ctx.rev()
todo = [rev]
while todo:
rev = todo.pop()
if rev in latesttags:
continue
ctx = repo[rev]
tags = [t for t in ctx.tags()
if (repo.tagtype(t) and repo.tagtype(t) != 'local')]
if tags:
latesttags[rev] = ctx.date()[0], 0, [t for t in sorted(tags)]
continue
try:
# The tuples are laid out so the right one can be found by
# comparison.
pdate, pdist, ptag = max(
latesttags[p.rev()] for p in ctx.parents())
except KeyError:
# Cache miss - recurse
todo.append(rev)
todo.extend(p.rev() for p in ctx.parents())
continue
latesttags[rev] = pdate, pdist + 1, ptag
return latesttags[rev]
def getrenamedfn(repo, endrev=None):
rcache = {}
if endrev is None:
endrev = len(repo)
def getrenamed(fn, rev):
'''looks up all renames for a file (up to endrev) the first
time the file is given. It indexes on the changerev and only
parses the manifest if linkrev != changerev.
Returns rename info for fn at changerev rev.'''
if fn not in rcache:
rcache[fn] = {}
fl = repo.file(fn)
for i in fl:
lr = fl.linkrev(i)
renamed = fl.renamed(fl.node(i))
rcache[fn][lr] = renamed
if lr >= endrev:
break
if rev in rcache[fn]:
return rcache[fn][rev]
# If linkrev != rev (i.e. rev not found in rcache) fallback to
# filectx logic.
try:
return repo[rev][fn].renamed()
except error.LookupError:
return None
return getrenamed
def showauthor(repo, ctx, templ, **args):
""":author: String. The unmodified author of the changeset."""
return ctx.user()
def showbisect(repo, ctx, templ, **args):
""":bisect: String. The changeset bisection status."""
return hbisect.label(repo, ctx.node())
def showbranch(**args):
""":branch: String. The name of the branch on which the changeset was
committed.
"""
return args['ctx'].branch()
def showbranches(**args):
""":branches: List of strings. The name of the branch on which the
changeset was committed. Will be empty if the branch name was
default.
"""
branch = args['ctx'].branch()
if branch != 'default':
return showlist('branch', [branch], plural='branches', **args)
return showlist('branch', [], plural='branches', **args)
def showbookmarks(**args):
""":bookmarks: List of strings. Any bookmarks associated with the
changeset. Also sets 'active', the name of the active bookmark.
"""
repo = args['ctx']._repo
bookmarks = args['ctx'].bookmarks()
active = repo._activebookmark
makemap = lambda v: {'bookmark': v, 'active': active, 'current': active}
f = _showlist('bookmark', bookmarks, **args)
return _hybrid(f, bookmarks, makemap, lambda x: x['bookmark'])
def showchildren(**args):
""":children: List of strings. The children of the changeset."""
ctx = args['ctx']
childrevs = ['%d:%s' % (cctx, cctx) for cctx in ctx.children()]
return showlist('children', childrevs, element='child', **args)
# Deprecated, but kept alive for help generation a purpose.
def showcurrentbookmark(**args):
""":currentbookmark: String. The active bookmark, if it is
associated with the changeset (DEPRECATED)"""
return showactivebookmark(**args)
def showactivebookmark(**args):
""":activebookmark: String. The active bookmark, if it is
associated with the changeset"""
active = args['repo']._activebookmark
if active and active in args['ctx'].bookmarks():
return active
return ''
def showdate(repo, ctx, templ, **args):
""":date: Date information. The date when the changeset was committed."""
return ctx.date()
def showdescription(repo, ctx, templ, **args):
""":desc: String. The text of the changeset description."""
return ctx.description().strip()
def showdiffstat(repo, ctx, templ, **args):
""":diffstat: String. Statistics of changes with the following format:
"modified files: +added/-removed lines"
"""
stats = patch.diffstatdata(util.iterlines(ctx.diff()))
maxname, maxtotal, adds, removes, binary = patch.diffstatsum(stats)
return '%s: +%s/-%s' % (len(stats), adds, removes)
def showextras(**args):
""":extras: List of dicts with key, value entries of the 'extras'
field of this changeset."""
extras = args['ctx'].extra()
extras = util.sortdict((k, extras[k]) for k in sorted(extras))
makemap = lambda k: {'key': k, 'value': extras[k]}
c = [makemap(k) for k in extras]
f = _showlist('extra', c, plural='extras', **args)
return _hybrid(f, extras, makemap,
lambda x: '%s=%s' % (x['key'], x['value']))
def showfileadds(**args):
""":file_adds: List of strings. Files added by this changeset."""
repo, ctx, revcache = args['repo'], args['ctx'], args['revcache']
return showlist('file_add', getfiles(repo, ctx, revcache)[1],
element='file', **args)
def showfilecopies(**args):
""":file_copies: List of strings. Files copied in this changeset with
their sources.
"""
cache, ctx = args['cache'], args['ctx']
copies = args['revcache'].get('copies')
if copies is None:
if 'getrenamed' not in cache:
cache['getrenamed'] = getrenamedfn(args['repo'])
copies = []
getrenamed = cache['getrenamed']
for fn in ctx.files():
rename = getrenamed(fn, ctx.rev())
if rename:
copies.append((fn, rename[0]))
copies = util.sortdict(copies)
makemap = lambda k: {'name': k, 'source': copies[k]}
c = [makemap(k) for k in copies]
f = _showlist('file_copy', c, plural='file_copies', **args)
return _hybrid(f, copies, makemap,
lambda x: '%s (%s)' % (x['name'], x['source']))
# showfilecopiesswitch() displays file copies only if copy records are
# provided before calling the templater, usually with a --copies
# command line switch.
def showfilecopiesswitch(**args):
""":file_copies_switch: List of strings. Like "file_copies" but displayed
only if the --copied switch is set.
"""
copies = args['revcache'].get('copies') or []
copies = util.sortdict(copies)
makemap = lambda k: {'name': k, 'source': copies[k]}
c = [makemap(k) for k in copies]
f = _showlist('file_copy', c, plural='file_copies', **args)
return _hybrid(f, copies, makemap,
lambda x: '%s (%s)' % (x['name'], x['source']))
def showfiledels(**args):
""":file_dels: List of strings. Files removed by this changeset."""
repo, ctx, revcache = args['repo'], args['ctx'], args['revcache']
return showlist('file_del', getfiles(repo, ctx, revcache)[2],
element='file', **args)
def showfilemods(**args):
""":file_mods: List of strings. Files modified by this changeset."""
repo, ctx, revcache = args['repo'], args['ctx'], args['revcache']
return showlist('file_mod', getfiles(repo, ctx, revcache)[0],
element='file', **args)
def showfiles(**args):
""":files: List of strings. All files modified, added, or removed by this
changeset.
"""
return showlist('file', args['ctx'].files(), **args)
def showlatesttag(**args):
""":latesttag: List of strings. The global tags on the most recent globally
tagged ancestor of this changeset.
"""
repo, ctx = args['repo'], args['ctx']
cache = args['cache']
latesttags = getlatesttags(repo, ctx, cache)[2]
return showlist('latesttag', latesttags, separator=':', **args)
def showlatesttagdistance(repo, ctx, templ, cache, **args):
""":latesttagdistance: Integer. Longest path to the latest tag."""
return getlatesttags(repo, ctx, cache)[1]
def showchangessincelatesttag(repo, ctx, templ, cache, **args):
""":changessincelatesttag: Integer. All ancestors not in the latest tag."""
latesttag = getlatesttags(repo, ctx, cache)[2][0]
offset = 0
revs = [ctx.rev()]
# The only() revset doesn't currently support wdir()
if ctx.rev() is None:
offset = 1
revs = [p.rev() for p in ctx.parents()]
return len(repo.revs('only(%ld, %s)', revs, latesttag)) + offset
def showmanifest(**args):
repo, ctx, templ = args['repo'], args['ctx'], args['templ']
mnode = ctx.manifestnode()
if mnode is None:
# just avoid crash, we might want to use the 'ff...' hash in future
return
args = args.copy()
args.update({'rev': repo.manifest.rev(mnode), 'node': hex(mnode)})
return templ('manifest', **args)
def shownode(repo, ctx, templ, **args):
""":node: String. The changeset identification hash, as a 40 hexadecimal
digit string.
"""
return ctx.hex()
def showp1rev(repo, ctx, templ, **args):
""":p1rev: Integer. The repository-local revision number of the changeset's
first parent, or -1 if the changeset has no parents."""
return ctx.p1().rev()
def showp2rev(repo, ctx, templ, **args):
""":p2rev: Integer. The repository-local revision number of the changeset's
second parent, or -1 if the changeset has no second parent."""
return ctx.p2().rev()
def showp1node(repo, ctx, templ, **args):
""":p1node: String. The identification hash of the changeset's first parent,
as a 40 digit hexadecimal string. If the changeset has no parents, all
digits are 0."""
return ctx.p1().hex()
def showp2node(repo, ctx, templ, **args):
""":p2node: String. The identification hash of the changeset's second
parent, as a 40 digit hexadecimal string. If the changeset has no second
parent, all digits are 0."""
return ctx.p2().hex()
def showphase(repo, ctx, templ, **args):
""":phase: String. The changeset phase name."""
return ctx.phasestr()
def showphaseidx(repo, ctx, templ, **args):
""":phaseidx: Integer. The changeset phase index."""
return ctx.phase()
def showrev(repo, ctx, templ, **args):
""":rev: Integer. The repository-local changeset revision number."""
return scmutil.intrev(ctx.rev())
def showsubrepos(**args):
""":subrepos: List of strings. Updated subrepositories in the changeset."""
ctx = args['ctx']
substate = ctx.substate
if not substate:
return showlist('subrepo', [], **args)
psubstate = ctx.parents()[0].substate or {}
subrepos = []
for sub in substate:
if sub not in psubstate or substate[sub] != psubstate[sub]:
subrepos.append(sub) # modified or newly added in ctx
for sub in psubstate:
if sub not in substate:
subrepos.append(sub) # removed in ctx
return showlist('subrepo', sorted(subrepos), **args)
def shownames(namespace, **args):
"""helper method to generate a template keyword for a namespace"""
ctx = args['ctx']
repo = ctx.repo()
ns = repo.names[namespace]
names = ns.names(repo, ctx.node())
return showlist(ns.templatename, names, plural=namespace, **args)
# don't remove "showtags" definition, even though namespaces will put
# a helper function for "tags" keyword into "keywords" map automatically,
# because online help text is built without namespaces initialization
def showtags(**args):
""":tags: List of strings. Any tags associated with the changeset."""
return shownames('tags', **args)
# keywords are callables like:
# fn(repo, ctx, templ, cache, revcache, **args)
# with:
# repo - current repository instance
# ctx - the changectx being displayed
# templ - the templater instance
# cache - a cache dictionary for the whole templater run
# revcache - a cache dictionary for the current revision
keywords = {
'activebookmark': showactivebookmark,
'author': showauthor,
'bisect': showbisect,
'branch': showbranch,
'branches': showbranches,
'bookmarks': showbookmarks,
'changessincelatesttag': showchangessincelatesttag,
'children': showchildren,
# currentbookmark is deprecated
'currentbookmark': showcurrentbookmark,
'date': showdate,
'desc': showdescription,
'diffstat': showdiffstat,
'extras': showextras,
'file_adds': showfileadds,
'file_copies': showfilecopies,
'file_copies_switch': showfilecopiesswitch,
'file_dels': showfiledels,
'file_mods': showfilemods,
'files': showfiles,
'latesttag': showlatesttag,
'latesttagdistance': showlatesttagdistance,
'manifest': showmanifest,
'node': shownode,
'p1rev': showp1rev,
'p1node': showp1node,
'p2rev': showp2rev,
'p2node': showp2node,
'phase': showphase,
'phaseidx': showphaseidx,
'rev': showrev,
'subrepos': showsubrepos,
'tags': showtags,
}
def _showparents(**args):
""":parents: List of strings. The parents of the changeset in "rev:node"
format. If the changeset has only one "natural" parent (the predecessor
revision) nothing is shown."""
pass
dockeywords = {
'parents': _showparents,
}
dockeywords.update(keywords)
del dockeywords['branches']
# tell hggettext to extract docstrings from these functions:
i18nfunctions = dockeywords.values()